My Review of Designing Data Visualizations

March 3, 2012

Originally submitted at O’Reilly

Intentional Communication from Data to Display

Data Visualization …Visualized Badly

By That Ontology Guy from Fredericton, NB on 3/2/2012

 

2out of 5

Pros: Easy to understand, Concise

Cons: Not comprehensive enough, Too basic

Best Uses: Student

First with the positives: It is a good introductory read into the subject. An average person could likely read in under an hour and take away some basics. (I read it on the bus on the way home). I would immediately recommend reading some of the many references listed for more detailed information. I did like the distinction between data visualizations that are largely one-off builds that almost require a project in themselves to create versus basic effective visualizations. Most people only have time for the basic visualizations. Missing was a description of how some people “jazz” up graphics bar charts by making them “3-D” but effectively making it too hard to compare the data.

Now the negatives: 1. I received this book as a reviewers copy in hard copy. I wish I would have had the PDF because the 32 pages of color graphics (for only a 93 page book) meant I immediately had to find a color printer to print them off so I could really understand them. The book size dimensions also really limited their size and thus their interpretability.

With the introductory nature of the subject coverage and the color demands and size demands of the figures, I feel this book would have made a great web-site. As it stands, it is a basic primer good for a school or public library.

(legalese)

UNB Computer Science Expo 2010

April 14, 2010

I just attended (13 April 2010) the UNB Computer Science Expo at the University of New Brunswick Fredericton.

It was a great Industry-University forum to discuss business development partnerships (Enterprise Fredericton, Propel ICT, NB IT Council), funding (New Brunswick Innovation Fund), human resource development (MITACS), example success stories (Chalk/RIM), as well as specific research projects of faculty and graduate students.

As an example of its utility, my current research at Innovatia is in Knowledge Engineering creating Ontology’s for text mining. At the expo, I found the following talks to be useful:

  • Ken Kents research in Accelerating Java Using Massive Multicore Systems to be useful for understanding how to overcome the large Triple Store we will face when operationalizing our RDF store (database).
  • Huajie Zhang research in Semi-supervised Learning and Its Applications to be useful for re-validating how grassroots users will be needed to help build the ontology for an organization.
  • Judy Zhao research in Managing Uncertain Knowledge on the Fuzzy Semantic Web to be useful for reminding me that one group of users can have precise queries, while other users (like the public) can have fuzzy queries. In fact, while she was talking, she helped my realize how I could solve a fuzzy ontology problem I had around one fuzzy object relationship that had 8 precise meanings. I am now modelling that fuzzy relationship with no description logic (axioms) but the sub-object properties that have precise meaning with axioms. This will allow us to suggest to the average user that we think their query is actually asking the one precise relationship meaning based on other aspects of the query (class/instance terms). Tomorrows job is to hammer out the details with the project team.
  • Nicola Bicocchi presentation on Collaborative Organizational Modelling and Simulation for Emergency-Response where he talked about 5 different models and how none interact well when responsinf to an emergency. I think ther is some OWL-2 ontology work here. Must call him in the morning.
  • And of course, Chris Baker’s presentation (On the Illustration of Ontology Reuse ) where he featured my ontology and knowledge engineering project plan. :-)
  • Alexandre Kouznetsov’s poster on  Text Mining & NLP based Algorithm to populate ontology with A-Box individuals and object properties where he thanked me for my ontology (THANKS ALEX!) and I got to talk about my work. :-)
  • And some interesting research posters: http://www.cs.unb.ca/research-expo/2010/2010Posters.htm
  • And, to pursue my PhD studies, I need to talk to Ian Baird at MITACs about my particular problem of trying to find funding for studies when I am already employed in Industry. Again, another phone call in the morning.

The great thing about this is that I discovered all this talent here in UNB when I have been looking at reasearch all over the world to help with our Semantic project. I had conversations with the NRC (William Langley) about how we could create networking between NB industry and UNB with a suggestion of www.academia.edu and Google Alerts playing a role. So I spent an hour writing him an email prior to this blog post instead of grading some early submitted term papers for RMC.

I also had some brief conversations with other industry participants about how semantics can help them. I think I convinced a few of them of its utility and hopefully some projects will emerge where I can use my ontology creation skills.

All in all a succesful day.

Reply to George Siemens: Using Technology to Improve Collaboration

December 14, 2009

Read with interest George Siemens post on creating simple information sharing environments http://www.elearnspace.org/blog/2009/12/12/using-technology-to-improve-collaboration/. I just designed a course for St Thomas University using open source tools like yourself for CCK08 and 09.

http://sts2413a2010.wikispaces.com/

I am using Google to make most of the functionality work ie Alerts, a Forum, and WordPress for blogging, posting papers. I went open because STU uses WebCT and is switching over to Moodle and I still have not been officially hired yet so I have little time to learn a new LMS.

This brings me to my second point of your point of simplicity over complexity. You note SharePoint problems. Yes it is complex. But it is a useful toolset when trying to organize complex business and making information findable. The portal aspect of SharePoint may be its greatest strength because it does not require much skill to learn to add sites, lists and libraries with standard navigation elements, thus making it easier to find things. I have found after assisting several organizations cleaning up their SharePoint implementation was that they continue to folder their records, treat SharePoint as a fileserver instead of a portal/intranet, and fail to apply metadata to make documents findable, sortable and viewable (ie usability and findability).

You note that I favour SharePoint for the business, but that is because I have seen the power of integrating Project Management, records management, document workflow and publishing, status reporting, and communications (SharePoint can even issue SMS messages!). At the same time, I can also go open source.

One aspect of my STU solution is no grade book. This is a weakness. Wouldn’t a student love to see a grade at the end of my commentary about their paper, rather than have to go to another system to see this confidential info. I have criticized Desire2Learns LMS because of the difficulty of assigning “grades” to discussion posts (RMC does a combined quality and quality metric for discussion forums).

Ultimatley, you have to decide what the information management problem is, what you want to get out of the solution, and what you can afford. Some days it may be SharePoiunt, other days it may by a Google solution.

Utility Of Twitter and a Networking Analogy

October 1, 2009

I replied to a philosophical posting about Twitter by BlanchManyard. Here is what I said:
I find myself being drawn more to Twitter for finding out about useful [online] reports from my “like-minded” network. I can’t always sign up for all the interesting RSS feeds or email notifications, so it is a useful complement. What is different than the neighbors analogy is that I am privy to conversations, or at least your side of conversation with a neighbor you are talking to over your back fence when I live across the road from you and thus would not tend to interact with your over-the-fence-neighbor. I also find out things about you that I would not normally hear from you becasue I can see everything you are saying 24-7, whereas if I was your neighbor, I might talk to once or twice a day. Twitter may be more analogous to the conversations you have with your spouse when you have 2 young children – burst transmissions of no more than 2 sentences before you get cut off or someone else interrupts.”

Memetics and Connectivism

September 20, 2009

Hello Connectivism & Connected Learning 2009,

I took the course last year. I just finished reading an article by Andrew Sullivan on memetics and how it applies to strategic communications to counter terrorism ideologies (his article). I immediately thought of how to relate this to Connectivism. The Sullivan talks about how the broadcast medium and the internet are the focus of counter-terrorism message sending, however, among the Islamic community, people do not listen to the message because the counter ideology to Islamic ideology is a head-on attack to the teaching of the Koran. Thus, no one listens to Voice of America etc for their counter-terrorism ideological message. Sullivan argues, that to be effective, strategic communications needs to occur at the grass roots levels.

In many cultures, our values, beliefs, and ideology is formed through those we have close contact with. Therefore, to be successful in Iraq and Afghanistan, NATO must take its message directly to the people. And counter terrorism message cannot be a direct assault on Islam. Instead it must be what he calls “oblique” or what the English Interwar period military strategist, JFC Fuller calls, the indirect approach to defeat your enemy.

So what does this talk of strategic communications and counterterrorism mean to connectivism? Well connectivism is partly about how the social web allows many disparate people from all over the word connect with those of similar interests. Sullivan argues that technology will have little impact, instead you have to get the “oblique ideas” into peoples heads through verbal and personal discussions. (I guess that is why the Provincial Reconstruction Teams are so important.) So, Connectivsm and connected learning do not appear to be valid in this case. Read Little Boxes, Glocalization and Networked Individualism for another viewpoint that may support Sullivan’s thesis.

However, in support of connectivism and connected learning, I would argue that the broadcast and internet communications have to be more subtle. Rather than be a “bulldozer” you have to be a “gardener” to cultivate change in ideas. This is a principle of Facebook and Twitter. Companies that try to force their message out there, do not do well. It is better to have a conversation. So NATO and the US would be well advised to take advantage of the social aspects of the web to counter terrorist ideology.

In Afghanistan, the context of communication is the personal dialog. Therefore, the message must be carried by real people. In Afghanistan, maybe Connected Learning is the personal. So the trick is to get the message out to as many areas as possible and that it be consistent, but local. Perhaps NATO’s use of information technology allows the messengers (the PRT team) to be rapidly responsive to changes and to disseminate that information to all the military forces, to the Afghan government and ultimately onto the Afghan people.

POE206 Fall 2009 and New Experiment

September 14, 2009

I am teaching POE 206, RMCs Canadian Civics course. I am doing it completely online compared to teaching it to a live class like I did in  Winter 2009, so students have some essay requirements, discussions in online forums, and some timed quizzes composed of an essay question.

I disagree with the time essay question on two points: 1. Timed testes especially essays, result in regurgitation and not critical analysis. 2. Many of the students are on operations or temporary duty whether home or abroad, and while the quiz can be done anytime during a specific week, they only have 2 hours. Some students will not have connectivity that week, thus forcing me to modify the quiz  time frame in the LMS or accept email submissions based on an honor system. How many students will this happen to? Well its been my experience at least every student has to ask for a modified submission date for at least one assignment per my courses. This makes at least 25 out of 175 assignment submissions have to have negotiated alternate arrangements. So why not go to an honor system like the History department at RMC does for HIE208 and HIE275?

Now for my experiment. In my Winter POE206 course  that I blogged about previous (Post 1, Post 2, Post 3, Post 4), I had the student stake a political orientation survey. I found the majority of the class was conservative leaning last winter, which surprised me because the Military was Liberal leaning in the 1990s. So I suspected in the Winter of 2009 that it had something to do with recent budget increases, purchases of new equipment, and what soldiers don’t love a war, and the conservatives are running one (Despite the fact it was the Liberals that took us int that war.)  So for my fall course, I have decided to ask the next question as ask the students to recollect who the voted for in Sept 2008, January 2006 and June 2004 and why they voted for who they did (local considerations , federal leader, party platform, did not vote, spoiled vote).

Whit this info, I expect to be able to answer the influence of the orientation survey.

The next thing that interests me is if this course will have an impact on how students view politics. Many students report disliking politics or often finding themselves ignorant of the Canadian system, so I wonder if the course will change their views. Thus, I plan on asking them in the last week to retake the political orientation survey to see if their scored change and in which areas that they do. If there is a forced election this fall, then I will have even better information to work with!

Perhaps I can turn this into a combined Masters of education and political science degree? perhaps I can compare how I was able to teach the course in Winter 2009 (live students in the class)  in the Winter of 2009 to Fall 2009 (online versions) to see which course is more effective by looking at grades and studnets qualitative comments to me?

A couple issues come up here: 1. Can I get into a MEd program quick enough to take advantage of the situation. 2. Will RMC get annoyed with me by having students be surveyed without their permission, 3. What ethical questions do I raise not only with my employer (RMC), but by using my own course as the basis of a MEd thesis when the students were taking the course for credit and I am probing them for information? (Do they answer the survey because they fell that to not answer the survey will affect their grade?)

I look forward to your comments.

Results of POE206 Mock Parliament

April 7, 2009

A quick update on the Mock Parliament I ran for my POE 206 Classs this term. My students said they enjoyed the exercise. However, they said they were prefer to be assigned to a political party during the role play that was not of their political orientation. In other words, during the survey at the beginning of the course, if they identified they were of Conservative leanings, they would prefer to have played the role of a Liberal or NDP member of parliament. From a pedagogical perspective, I ahve to agree with them. They would not only learn about their own poltical orientation, but also the opposite point of view. The students also suggested arranging 2 short half hour prepatory sessions to get organized for the Mock Parliament. They felt the online discussion forum I ahd set up was insufficent for real time discussion. Again I ahd to agree. So much of politics is about real time interpersonal communication.
From my perspective, I thought it went well although there was some administrative glitches:
1. The students presented a bill that no-one had actually addressed as an issue.
2. The governing party did not prepare hardcopy versions of their bill, something Parliament requires.
And just like in the real Parliament, the oppositon had a lot to say about the bill but on the governemtn side, repsonse came from a few “leaders”. The back benchers were quite like the real world. The government in this case was Conservative and they kept true to Stephen HArper’s government stance.
Now I am waiting for the students Concept Maps and Final reports based on the issue area they chsoe to discuss this term.

POE206 Update: Mock Parliment

February 27, 2009

I wanted to give a brief update about my “connected” POE 206 class now the day 1 of 3 in the Mock Parliament has happened. We spent the last 3 weeks having students give presentations on political subjects of their choice such as education reform, First Nation land claims, Sikh right to not wear helmets while driving a motorcycles and so on. When the last presentation was given this week on education reform, we had a big discussion on the 3 functions of industrialized education:

  • create productive workers,
  • create a national unity,
  • create a civic minded population.

When the presentation was over, I had the students split into the government and opposition for the mock trial based on their responses to a political orientation survey I gave. As soon as the class split into its Liberal, NDP, and Conservative party groups, one lady commented that now she understood why certain people had certain views on political issues. Once it became clear what their party affiliation was, their commentary made sense. Last night I had 11 conservatives, 4 Liberals and 1 NDP. I had noticed that the class had a pro-conservative bias from the very first night. This was different that the military I knew in the 1980-1990s which was much more evenly split between Liberals and Conservatives.

I then wondered how perhaps the Afghanistan mission may have changed military members political orientation and posed the question:

Who of each party had voted for another party in 2004 other than the one they were in last night.

It turns out 4 of 11 Conservatives had voted Liberal in 2004. And 2 of 4 liberals had voted conservative. My NDP had voted NDP in 2004. Looking a the numbers then, I had 7 Liberals/NDP in 2004 and 9 Conservatives, a much more even split. I then stated to the class  that even though the Conservatives were in power, the Liberals had committed the Armed Forces to Afghanistan. The Armed Forces take great pride in their mission in Afghanistan and view it as a cathartic event because for the first time since the Koren War, the military has a clear purpose. The Canadian Armed Forces is reaching within and re-discovering its reason d’etre: “To win the nations wars.”

Soon after going to Afghanistan, a national election gave the Conservative a minority government. Since then, they have had to spend significant sums of money on new equipment for the mission or to support the Canadian Forces in general: C-17 strategic air lifter, CH-47 medium lift helicopter, Cormorant Anti-submarine helicopter, Leopard 2 tanks, mine detection/clearing equipment, mine resistant vehicles, , unarmed aerial vehicles for reconnaissance, new artillery rounds, a satellite surveillance system to monitor the Arctic, with a new C-130 replace soon to come. With all this new equipment, why wouldn’t a CF member support the current government . The Conservative government is making up for years of equipment declines. This is no different than any employee in a large organization pledging their support to a division or department that is currently undergoing expansion or growth.
So that is a political theory I am advancing on why the CF is currenlty very supportive of the conservative government.

CCK08: Course Critique for Connectivism and Connected Learning

February 21, 2009

I must admire George Siemens ad Stephen Downes for attempting the CCKo8 Connectivism and Connected Learning course in the Fall of 2008. The certainly bit off quite a bit trying to run an on-line course using many free software platforms and with 2200 users registered (although only about 30 paying students including yours truly). I have since created a online course, not the extent of CCK08, but I have seen my instructor hours go up by about 50% as I prepare presentations, deliver monitor discussion groups and mark more lessor valued assignments. (See my description of my POE 206 course for a comparison). My critique is largely based on the technical and administration. I actually found the content to be appropriate, stimulating, and thought provoking. I also went out and found extra things to read and since I subscribed to Stephen Downes OL Daily, I found myself reading other pertinent material not mentioned as required or optional reading in the Course Wiki. I do hope my comments below are taken in light of the fact that as a Canadian Military Historian by training and a former Army officer, I have a tendency to focus on the 5% that went wrong and not the 95% that went well :). I have an annoying tendency to strive for perfection.

George and Stephen set up a course that used Moodle for self contained discussions and posing of questions, a Wiki for the outline, A course daily newsletter, several blog platforms for students blogging, Elluminate and Ustream for live discussion, and tied the various tools together using Google alerts to notify of blog posts. I personally found the proliferation of platforms to be a nuisance from a technical and attention span perspective. However, it did prove the point of the how various networks operate and how different people communicate. Each platform addresses a particular user group and their preferences. I have been somewhat spoiled by being able to use  Enterprise grade Learning Management Systems (Desire2Leanrn) and Intranet platforms (SharePoint 2007). Also, George did warn us we would not be able to follow all activities (I skipped Second Life).

On the technical side, I could not access the Wiki until my company permitted the University of Manitoba port for the site through as the U of Manitoba used an non-standard. I was never able to get the chat feature in UStream to work, thus I missed out on a lot of back channel discussion during the Friday meetings. UStream also failed 3 times and we had to revert to Elluminate. There was no easy way notify the students of this. I would suggest in the future that students are told to monitor Twitter an hour before the class and if the session is cancelled (it happened once or there are problems with a platform, that the students could switch to the alternate platform.) George told me that some students could not use Elluminate and I was the only one that could not use Ustream properly so there was a trade off. Unfortunately, my teaching philosophy is that there has to be 100% access on the part of the students otherwise you cannot use that technology (I run into this problem all the time with the military DWAN network not allowing the use of Active X components or the download of Microsoft Office products. Thus, I cannot link to You tube, CBC videos, and I must post all my files as PDFs.) Since George is now aware of these technical issues, perhaps they can be investigated, solutions found, and int he future students can be advised of what system requirements, settings, and network security permissions they may need. I also found the need to go in a subscribe to the weekly Moodle forums to be a nuisance. I think Moodle needs an option to allow a person to automatically subscribe to all forums as they are created. In my own LMS, all weekly forums are present before the first students ever comes in so this is not an issue. Perhaps George could create each weeks forum in advances. This would also cause the students to post their comments in the appropriate week pertain to the subject of their comment as opposed  posting early.

As for attention span, the multitude of platforms made it difficult to track what was going on, especially blogs. As such, I found myself monitoring mostly e Moodle Discussion forum and using Google Alerts and Stephen and Georges noteworthy blog mentions to monitor blogs. I never did track any one else blog other than Lisa Lane because I find monitoring RSS feeds cumbersome (I am a power Email user). I am sure George and Stephen will have some obvious advice for me on that :) I did notice the George did have difficulty tracking when people had submitted assignments because they could appear in your personal blog, Moodle or on other platforms. George had to rely on tracking the paying students blogs, Google Alerts and watching for tell tale subject lines t tan assignment was posted. I think what needs to happen for assignments that are to be graded is that there needs to be a clear subject line like”CCK08 Assignment 1 Submission: then your catchy title”. This would make it clear to all. Another hint would be to require students to send an email to George or require them to make a posting in a Moodle Assignment Submission Forum and provide the link to the site. I myself have found this to be a problem in POE 206  so I developed clear file name conventions, and subject line headings so general postings did not cause me to lose sight of the gradable assignments.

The last item is the delay in being notified of my final grade. Unfortunately, U of Manitoba has to have all students grades in before they can release it to individuals. I paid for the course out of my pocket and my company is reimbursing me upon receipt of a passing grade. I was starting to miss the $395 and my company wanted to finalize its 2008 financials. When I questioned George about this in the first week of February, he said he was still waiting for some final assignments. So, U of Manitoba should change its policy on releasing grades upon individual completion instead of class completion. For example, RMC uses an LMS that allows me to enter grades as the course progresses. As soon as the final grade for a person comes in, I enter it, and then “release” the grade the student can see their final grade immediately. I do have students that arrange to submit late, but there is a formal administrative notification and tracking system for late submitters and one or 2 people do not hold up the rest of the class.

I have 2 other suggestions:

1. Dave Cormier did an interesting experiment with UStream one week in which he ran 3 videos segments at the same time (himself, George, and Stephen).I thin this should be explored some more as I would have found it interesting to see all 24-50 students streaming at the same time (Although my network administrator would have freaked out at the bandwidth consumed.) Is there a live web streaming aggregator platform that could do this for only modest bandwidth usage. Just think how useful that could be, kind of like watching CNN, Fox, ABC, NBC, CBS, CBC, BBC etc all at once. This obviously has commercial applications for remote workers or even the extended family int he general public.

2. The Wiki Course Outline, although open for all to contribute to, was largely left untouched. Perhaps greater use could be made of this or perhaps a second wiki page to allow interesting student contributions to each subject in the course outline.

I would also suggest someone do a network study on this course as mentioned by several students in the Moodle. George admitted it would be hard to get permission from the university. I discovered some marketing monitoring software developed by Radian6 in Fredericton NB has the ability to go back 6 months and look at the internet historically to see who is talking about CCK08 and what they are linking to.

Despite 3 long paragraphs of what are largely “technical issues” and no philosophical or pedagogical issues, I think version 2.0 of the course will likely run a lot smoother technically, and require less time on the part of the facilitators. I enjoyed the course experience and hope to take more courses leading up to the new Certificate program at U of Manitoba in Certificate in Interdisciplinary Studies: Emerging Technologies for Learning (ETL)

Signing off of CKK08 (although I will  keep my Google Alerts active.)

POE 206 Canadian Political System Concept Map

February 18, 2009

My students did up concept maps of the Canadian Political system for the Canadian Forces professional development course as an assignment. Some had difficulty created “network connections” between ideas so I did one up to guide them further as they refine theirs as the course moves into the second half. Here is mine (by the way, the colouration of the political parties in the map is the colours the parties use when election campaigning):

canadian-political-system-what-is-the-basis-of-the-canadian-political-system


Follow

Get every new post delivered to your Inbox.